calibre/recipes/la_razon_bo.recipe

65 lines
3.0 KiB
Plaintext

__license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
'''
www.la-razon.com
'''
from calibre import strftime
from calibre.web.feeds.news import BasicNewsRecipe
class LaRazon_Bol(BasicNewsRecipe):
title = u'La Razón - Bolivia'
__author__ = 'Darko Miletic'
description = 'El diario nacional de Bolivia'
publisher = 'Praxsis S.R.L.'
category = 'news, politics, Bolivia'
oldest_article = 1
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'cp1252'
use_embedded_content = False
language = 'es_BO'
publication_type = 'newspaper'
delay = 1
remove_empty_feeds = True
cover_url = strftime('http://www.la-razon.com/portadas/%Y%m%d_LaRazon.jpg')
masthead_url = 'http://www.la-razon.com/imagenes/logo.jpg'
extra_css = """ body{font-family: Arial,Helvetica,sans-serif }
img{margin-bottom: 0.4em}
.noticia-titulo{font-family: Georgia,"Times New Roman",Times,serif}
.lead{font-weight: bold; font-size: 0.8em}
"""
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
keep_only_tags = [dict(name='div', attrs={'class':['noticia-titulo','noticia-desarrollo']})]
remove_tags = [dict(name=['meta','link','form','iframe','embed','object'])]
remove_attributes = ['width','height']
feeds = [
(u'Editorial' , u'http://www.la-razon.com/rss_editorial.php' )
,(u'Opinión' , u'http://www.la-razon.com/rss_opinion.php' )
,(u'Nacional' , u'http://www.la-razon.com/rss_nacional.php' )
,(u'Economia' , u'http://www.la-razon.com/rss_economia.php' )
,(u'Ciudades' , u'http://www.la-razon.com/rss_ciudades.php' )
,(u'Sociedad' , u'http://www.la-razon.com/rss_sociedad.php' )
,(u'Mundo' , u'http://www.la-razon.com/rss_sociedad.php' )
,(u'La Revista' , u'http://www.la-razon.com/rss_larevista.php' )
,(u'Sociales' , u'http://www.la-razon.com/rss_sociales.php' )
,(u'Mia' , u'http://www.la-razon.com/rss_mia.php' )
,(u'Marcas' , u'http://www.la-razon.com/rss_marcas.php' )
,(u'Escape' , u'http://www.la-razon.com/rss_escape.php' )
,(u'El Financiero' , u'http://www.la-razon.com/rss_financiero.php')
,(u'Tendencias' , u'http://www.la-razon.com/rss_tendencias.php')
]
def preprocess_html(self, soup):
for item in soup.findAll(style=True):
del item['style']
return soup